Data citation and reuse practice in biodiversity - Challenges of adopting a standard citation model

Abstract

© 2015 American Institute of Physics Inc.. All rights reserved. Openly available research data promotes reproducibility in science and results in higher citation rates for articles published with data in biological and social sciences. Even though biodiversity is one of the fields where data is frequently reused, information about how data is reused and cited is not often openly accessible from research data repositories. This study explores data citation and reuse practices in biodiversity by using openly available metadata for 43,802 datasets indexed in the Global Biodiversity Information Facility (GBIF). Quantitative analysis of dataset types and citation counts suggests that the number of studies making use of openly available biodiversity data has been increasing in a steady manner. Citation rates vary for different types of datasets based on the quality of data, and similarly to articles, it takes 2-3 years to accrue most citations for datasets. Content analysis of a random sample of unique citing articles (n=101) for 437 cited datasets in a random sample of 1000 datasets suggests that best practice for data citation is yet to be established. 26.7% of articles are mentioned the dataset in references, 12.9% are mentioned in data access statements in addition to the methods section, and only 2% are mentioned in all three sections, which is important for automatic extraction of citation information. Citation practice was inconsistent especially when a large number of subsets (12—50) were used. This calls for adoption of a standard citation model for this field to provide proper attribution when using subsets of data.